MSCI 223 Business Modelling and Simulation 2022
Coursework
Deadline for handing in this case: Friday April 22, 2022 at 10.00
This is the submission deadline and any submission after this deadline is subject to standard departmental penalties, unless you have been given an extension for exceptional reasons. The extension must be granted by me before the deadline.
Copyright By PowCoder代写 加微信 powcoder
Teamwork! I suggest you NOT to divide questions among group members as these questions ask different aspects of one simulation. You all need to participate in building the simulation. In your submission ZIP-file, please also include a one-page statement on each member’s contribution in the coursework.
You are requested to make the following assignments. You are allowed to work in teams of 5 – 7 students. Please, take the following into consideration when handing in this case:
1. Save each level under a different name (‘level1.mod’, ‘level2.mod’, …). You must hand in
all levels separately!
2. Regularly use the “save” button to prevent losing your work when the computer or software
crashes. When you are finished with level 1 and want to continue with level 2, then first
save your current model as “level2.mod” and only make changes after that!
3. Use the pre-structured form to hand in your results and your answers to the questions. You can find this form on Moodle under “Coursework”. Please stay within 8 pages (font size
4. Your models (one for each level), as well as the answer form, must be handed in via
Moodle. Please do create the ZIP-file precisely as instructed. Create a ZIP-file containing the files level1A.mod, Level1B.mod, Level2.mod, Question6.mod, caseanswerform.doc, and contributionstatement.doc. Put your group number in the name of the ZIP-file. So, for example, the name of the ZIP-file becomes “group 1.zip”. Please do not use any other compression software (such as RAR or ARJ).
5. You can obtain support for the level 1 assignments from me at the workshop or during the office hour. No support is given for answering the theoretical questions or for level 2.
6. The level 2 assignments generally require more advanced modelling skills that sometimes
have not (yet) been treated in the lectures. It may help to study the slides, other parts of
your Witness guide and examples of upcoming lectures.
7. You can obtain at most 70 points (distributed evenly over six questions and your simulation files) by making the level 1 assignment and answering the theoretical questions. Level 2 gives you at most 35 points (distributed evenly over one question and your simulation file). The maximum grade is 100.
8. There is a special grading arrangement for the waiting-line problem at the operating room. There are several aspects to consider. It is part of level 1A to make sure that no more than 1 patient can ever be in the queue. There are more issues though (using the standard solution you may send away too many patients). For all teams together, there are 2 extra points available for those who find and solve all issues. So if only 2 teams manage to solve this, they both receive 1 point extra. If 4 teams solve it, each of these teams gets 0.5 points extra. And so on. If you wish to participate in this, then (1) hand in your “base case” model as level1A.mod and (2) hand in the model with the expanded expression as level1Aextra.mod.
case – page 1
Emergency Room Simulation
As a renowned simulation expert you have been asked by a hospital to perform a study at its Emergency Room. This is usually a hectic environment. Large numbers of patients arrive and each time a decision has to be taken concerning treatment. Is a major surgery inevitable or will a plaster do? The purpose of this study is to support the management in their decision making. Because the hospital management consists of doctors who have grown into management, they seem to be unable to explain clearly what their ideas and targets are. You have been contacted because next year’s government funding will be lowered significantly. The only thing they know for sure is that waiting times should be limited to avoid that patients die or that their symptoms exacerbate.
First, you need to formulate a problem definition which you could present to the hospital’s management. Thereafter, you are advised to create a conceptual model and to supplement this subsequently with quantitative data to obtain a quantitative model. The models can be developed in a number of phases by gradually decreasing the abstraction level.
System description
Have a look at the map of the emergency room which is given to you by the management (case- map.doc). Patients arrive by car at the parking lot. On average one patient arrives every 3.5 minutes, according to an exponential distribution. Patients walk from the parking lot, through the entrance to a waiting area near the reception, which requires between 2 and 8 minutes (uniformly distributed). The patient sits down on a chair in the waiting area until it is his/her turn at the reception. The receptionist determines the appropriate specialisation based upon the nature of the injury. This takes on average 2 minutes, normally distributed with a standard deviation of 0.5 minutes. There are three possibilities:
1. Major injuries; these require an immediate complex operation elsewhere in the hospital. This concerns 5% of the patients. They go straight to another part of the hospital (out of the model).
2. Medium injuries; these require immediate operation; however, this can be performed at the Emergency Room. This concerns 7% of the patients. They go to one of the operating rooms where they will be treated by a surgeon. If all surgeons are busy, the patient must wait. The estimated time required to perform an operation is normally distributed with an expected value of 30 minutes and a standard deviation of 6 minutes. As you may have noticed, there is only one chair for patients who wait for an operation. After all, if the waiting line would increase, then the injuries may become fatal. Assume that the receptionist has sufficient information to decide whether to send a patient to the operating rooms or – if necessary to prevent a waiting line length of two patients – to send that patient to another part of the hospital (out of the model).
3. All other patients are sent to a doctor in one of the treatment rooms. The time required for a treatment is estimated as normally distributed with an expected value of 3 minutes and a variance of 0.25 minutes2. The doctor, who works from 9:00 until 17:00, is fairly inexperienced. Therefore, (s)he is 30% slower than his/her colleagues. The standard deviation remains the same.
Patients walk from the reception to their appropriate destination. It takes a patient about 2 minutes to leave the reception area (and the model) when going to another part of the hospital. To go from the reception to (the waiting room of) the treatment rooms takes between 3 and 4 minutes. To go from the reception to the operating rooms takes between 4 and 5 minutes.
A patient who has been treated by a surgeon has a 75% probability that (s)he can go home instantly after the operation. This patient then walks to the exit in about 4 minutes. Otherwise, the patient must recover in another part of the hospital. These patients walk after the operation to the door that leads to the other parts of the hospital in about 12 minutes. Patients who were
case – page 2
treated by a doctor can always go home after the treatment. It takes these patients about 4 minutes to walk to the exit.
There are 4 surgeons, 4 doctors and 3 receptionists. The receptionists work for 8 hours each (including breaks), from 7.00 until 15.00, from 15.00 until 23.00 and from 23.00 until 7.00. The receptionists take a 15-minute break every even hour. So the receptionist who starts at 7.00 takes breaks from 8.00 until 8.15, from 10.00 until 10.15, and so on. The surgeons work as follows. Three surgeons work in shifts, from 6.00 until 14.00, from 14.00 until 22.00 and from 22.00 until 6.00. The fourth surgeon works from 9.00 until 17.00. Each surgeon takes a 15- minute break every odd hour. So the breaks of the surgeon who starts at 6.00 are from 7.00 until 7.15, from 9.00 until 9.15, and so on. The fourth surgeon starts to work at 9.00 and therefore starts his/her day by taking a break. The schedules of the four doctors are identical to the schedules of the surgeons. Note that this is an Emergency Room, so if it is time for somebody to take a break, that person will first finish the patient (s)he is working on, before taking the break. Furthermore, the break will end at the scheduled time anyway.
Problem definition
Question 1
Formulate a problem definition (step 1 of a simulation project) which you could present to the hospital’s management.
Conceptual model
Question 2
Create a conceptual model of the process as described so far. Add quantitative data to obtain a quantitative model.
Implementation, Verification, Validation
Create a Witness model from the quantitative model you made at question 2. Simulate the system for 5 days. Start the simulation at 0.00.
Use the supplied map of the Emergency Room in your Witness model. Also think of the following in the animation of your model:
• The routes that patient walk along,
• It must be visible when a receptionist, doctor, or surgeon takes a break.
Furthermore, you get the following hints to make your model:
• Build the model gradually and check the functionality of each extension before you continue building.
• Model the surgeons and doctors with SETS. That will save you a lot of work at level 2.
• Think carefully about the SCHEDULES before you enter them into Witness. You only need 3 schedules, each schedule consisting of just a few (at most 9) ROWS.
• Check very very carefully (lives are at stake!) that there can never be more than 1 patient in the waiting line for the operating rooms. Also prevent sending patients to another part of the hospital if they would not cause the waiting line length to exceed 1. Examine the related processes carefully and find out all options that might occur. Before implementing, it might be wise to first write down a logic formulation that includes all of these options.
case – page 3
Question 3
Use the model from level 1 and run it for 5 days, starting at 0.00 to obtain the following results:
• The utilisation of the surgeons who work in shifts.
• The utilisation of the surgeon who works during daytime.
• The utilisation of the doctors who work in shifts.
• The utilisation of the doctor who works during daytime.
• The average waiting time at the treatment and operating rooms.
You discover that patients in reality do not arrive evenly during the day. From 7:00 until 19:00 one patient arrives every 2.4 minutes on average. From 19:00 until 7.00 one patient arrives every 6 minutes on average. Both distributions can be assumed to be exponential. Include this in the Witness level 1A model and save it as “level1B.mod”.
Question 4
What are the effects of this increased validity of the model on your results? Use again one run of 5 days for your comparisons (after all, we are not yet in the stage of precise output analysis).
Experiments and Output analysis
Question 5
The management of the hospital has provided you with a more detailed specification of their objectives. On average, patients who need to visit a doctor should need only 35 minutes from arriving at the parking lot until leaving the hospital. Each of the patients who require surgery should leave the hospital or arrive at another part of the hospital on average within 1 hour and 10 minutes after arriving at the parking lot.
Can these criteria be met by the current process? Cleary explain the steps you have performed to obtain your answers.
If not, can you explain to the management what is causing the problems? Can you propose some solutions to the management (other than the one mentioned in Question 6)? The proposed solution do not have to be modelled in Witness.
Question 6
The management of the hospital has decided to replace the inexperienced doctor who works from 9.00-17.00 by a more experienced doctor who has the same working speed as the other doctors. Is there a significant change in throughput time of patients who need to visit a doctor? How did you come to this conclusion? Clearly explain the steps you have performed to obtain your answers.
In addition to your written explanation, please save your model of Level 1B as “Question6.mod” with the following settings:
• Make sure that the treatment process is adapted such that all doctors have the same working speed.
• Set the replication parameters in RUN- SETUP… to the values you have used to perform this analysis.
case – page 4
Furthermore
• Make a screen dump of the screen you have used to derive your results (you can do this
by pressing the PrtScr or PrintScreen button on your keyboard, then press Ctrl-V in MS Word).
Level 1 still contains a number of simplifications compared to the real system. Change your model according to the information given below to increase the validity of your model. These extensions can all be modelled in the same file. Continue with the model resulting from “level 1b” and save it as “level2.mod”.
a. In practice, the surgeons sometimes take over work from a doctor. However a surgeon will only take over work from a doctor if (1) all doctors are busy treating a patient and (2) there are no patients waiting for an operation. The procedure for the patients mods not change. Only now they may be asked to enter an operating room in stead of a treatment room. (Hint: you do not need to add extra modules to your model.)
b. About 20% of the patients who have been treated by a doctor need crutches before they can leave the hospital. There is a stock of crutches available near the treatment rooms. Stock levels are kept in a computer system. The stock is updated directly when crutches are given to a patient. If there are no crutches available for a patient, the order is still entered into the computer system, due to which the stock level will be displayed with a negative value. The medical supplies lending office checks the computer system once per hour (say, at the hour, so at 0:00, 1:00, 2:00 and so on) to see if the treatment rooms ran out of stock. If a stock-out is observed (zero or negative stock level), 5 sets of new crutches will be delivered to the treatment rooms, which requires just 5 minutes. The patients waits in the hallway in case of a stock out, which assures that the doctor can continue his work on the next patient.
c. The management has complained that your model is invalid. They have noticed that patients seem to jump right out of the waiting room chair into the treatment rooms (they didn’t see that this actually also holds for the operating rooms). The management expects patients to take about 1 minute to walk this distance. Update your model to convince the management.
Question 7
Do the changes made at level 2 have an impact on the results?
case – page 5
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com